Age Differences in Spatial Cognition and Navigation: A Comparative Neuropsychology Approach

Material Information

Age Differences in Spatial Cognition and Navigation: A Comparative Neuropsychology Approach
KING, EMILY GREEN ( Author, Primary )
Copyright Date:


Subjects / Keywords:
Alzheimers disease ( jstor )
Cognition ( jstor )
Hippocampus ( jstor )
Maps ( jstor )
Memory ( jstor )
Navigation ( jstor )
Older adults ( jstor )
Performance spaces ( jstor )
Psychometrics ( jstor )
Young adults ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Emily Green King. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:


This item is only available as the following downloads:

Full Text




Copyright 2006 by Emily Green King


iii ACKNOWLEDGMENTS I would like to thank Dr. Russell Bauer fo r his guidance and mentorship throughout this project and my gradua te training; Dr. Kevin Thom as for his dedication and encouragement; my committee, Drs. Sheila Eyeb erg, Bill Perlstein, a nd Michael Perri, for their contributions to this document; and Yu -Ling Chang for her collaboration. Finally, I would like to thank my husband, Adrian, and my family for their love and support.


iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF LIST OF FIGURES..........................................................................................................vii ii CHAPTER 1 INTRODUCTION........................................................................................................1 Preclinical Detection of AlzheimerÂ’s Disease..............................................................1 The Medial Temporal Lobe Memory System (MTMS)...............................................4 Mild Cognitive Impairment (MCI)...............................................................................5 Cognitive Mapping Theory and the Role of the Hippocampus in Spatial Mapping....7 Testing Allocentric Spatial Memory and the Morris Water Maze (MWM).................8 Comparative Neuropsychology....................................................................................9 Further Assessment of Spatial Abilities in Aging Populations..................................10 Specific Aim 1............................................................................................................14 Specific Aim 2............................................................................................................15 Specific Aim 3:...........................................................................................................16 2 METHODS.................................................................................................................17 Participants.................................................................................................................17 Procedures...................................................................................................................18 Neuropsychological Screening and Assessment Battery....................................19 Spatial Cognition Psychometric Battery.............................................................21 Computer-Generated Arena.................................................................................22 Florida Route Learning Task...............................................................................25 Self-Report Measures of E nvironmental Spatial Ability.....................................26 Statistical Analyses.....................................................................................................26 3 RESULTS...................................................................................................................30 CG Arena Performance...............................................................................................30 Correlations Between CG Arena and Clinical Measures...........................................33


v Relationship Between Navigation in CG Arena Space and Real Space.....................34 Sex Differences...........................................................................................................35 4 DISCUSSION.............................................................................................................37 LIST OF REFERENCES...................................................................................................44 BIOGRAPHICAL SKETCH.............................................................................................52


vi LIST OF TABLES Table page 2-1 Demographic Characteristics of Ex perimental Participants by Group....................18 2-2 Neuropsychological and C linical Screening Measures............................................20 2-3 Spatial Cognition Psychometric Battery..................................................................21 3-1 Mean Performance (SD) on the C ognitive Tests as a Function of Age...................33 3-2 Correlations between CG Arena Composite and Clinical Measures.......................34 3-3 Correlations between CG Arena and FRLT.............................................................34 3-4 Mean Performance (SD) on the Cogn itive Tests as a Function of Gender..............36


vii LIST OF FIGURES Figure page 2-1 Representations of the C-G experiment al room as seen by the participants............24 2-2 Representation of target on ce it is successfully acquired.........................................24 2-3 Schematic representation of the Florida Route Learning Task (FRLT)...................26 3-1 Mean SE (± 1) path length deviation ratio of youn g and old participants over the eight acquisition trials..............................................................................................31 3-2 Mean SE (± 1) target crossings of young and old participants.................................31 3-3 Mean SE (± 1) percentage of time spent in th e northwest quadrant on the probe trial.......................................................................................................................... .32 3-4 Mean SE (± 1) CG Arena composite performance by group...................................32 3-5 Mean Florida Route Learning Task performance by group.....................................35


viii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science AGE DIFFERENCES IN SPATIAL COGNITION AND NAVIGATION: A COMPARATIVE NEUROPSYCHOLOGY APPROACH By Emily Green King May 2006 Chair: Russell M. Bauer Major Department: Clini cal and Health Psychology Extensive animal research has impli cated the hippocampus/medial temporal memory system (MTMS) in spatial mapping an d navigation, and pat hological studies of Alzheimer’s disease in humans have revealed that this system is among the first affected by the neuropathological changes th at are the hallmarks of the dis ease. It is in this context that we have been developing methods for identifying individuals who show signs of compromise in the MTMS. Elderly adults ar e more likely to report “getting lost” or becoming disoriented or confused as they na vigate their environment, and a loss of the ability to get around can result in considerab le loss of functional independence. More recently, the use of virtual navigation para digms has demonstrated the correlation of performance with cognitive map-based way-fi nding abilities. We exam ined these abilities in younger and older adults, and investigated the relationship among self-reported spatial ability/impairment, psychometrically defined spatial abilities, real world navigation, and spatial navigation in a computer-generated environment modeled after the Morris water


ix maze (the CG Arena). Twelve normal young adults (age 18-25 years) and 12 normal older (age 65 years and older) adults t ook part in this study. We first sought to characterize spatial naviga tion differences between young and old healthy adults. Consistent with previous lit erature and our hypothesis, young adults located the invisible target in shorter distance, more often, and sp ent a greater proportion of their total time in proximity to the location of the removed targ et on the probe trial of the CG Arena. Additionally, we sought to examine the relatio nship between current measures of spatial ability and CG Arena. Consistent with rese arch to date, few correlations were seen between CG Arena and psychometric or self-r eport measures of sp atial ability. Finally, we sought to examine the relationship between navigation in real spa ce and navigation in CG Arena space. We found that a real life navigation task was significantly correlated with CG Arena performance, but these correl ations were somewhat lower in magnitude when young and older adults were examined se parately. Taken collectively, these results suggest that navigation measures derived from the CG Arena capture unique variance that is not accounted for by self-report, psychometri cally defined measures of spatial ability, or real world experimental tas k. These data are consistent with theories that aging impairs the formation/retrieval of spatial maps of novel environments and spatial knowledge acquired from direct experience in the envi ronment, and with the notion that spatial mapping is a unique ability. If confirmed in a larger clinically defined mild cognitively impaired (MCI) sample, our preliminary findings suggest that virtual navigation measures are sensitive to real-world spatial navigation/orientation and have implications for predictive validity in the clinical setting.


1 CHAPTER 1 INTRODUCTION AlzheimerÂ’s disease (AD) is the most co mmon form of dementia in older adults, resulting in disability, dependency, and eventu al death from associated medical illnesses (Morris, 2005). As the population ages, there is growing concern that AD will become an enormous health problem with far-reaching social and economic consequences. According to recent census data, the prev alence of AD in the United States is approximately 4.5 million people (US Census Bu reau, 2000). It is predicted that this number will increase to 13.2 million by 2050 (Hebert et al., 2003), meaning that approximately 1 in 45 Americans will be afflicted with the disease at that point (Brookmeyer et al., 1998). As the oldest age groups in the United States, continue to grow, so will the prevalence of AD and th e proportion of severe cases. Due to the predicted epidemic proportions that AD is e xpected to reach worldwide, there is an increasing amount of research focusing on the prevention of AD as a means for treatment (Knopman, 2003). Preclinical Detection of AlzheimerÂ’s Disease In recent years, AD has become better unde rstood and several pot ential risk factors (e.g., age, family history, and th e presence of the apolipoprotein E 4 allele) have been identified (Morris, 2005). While AD can onl y be diagnosed conclusively at autopsy, several clinical methods allow for diagnosis of the disease with high accuracy, and many drugs have been approved to allow for the in itiation of treatment. Acetylcholinesterase inhibitors (ChEIs; e.g. donep ezil, rivastigmine, and galantamine) are recommended as the


2 standard treatment, but their effectiveness in patients with mild to moderate AD has been challenged (for a review, see Morris, 2005). In addition, antioxidants, anti-inflammatory agents, ampakines, and nootropics are being studied as potential AD therapies (for a review, see Petersen, 2003). The challenge is that, in order fo r these treatment options to prove to be truly effective, they need to be administered to at-risk individuals before the symptoms, and the underlying pathology appear. Because of this, understanding the tran sition from normal to pathological aging/dementia is critical. Mult i-disciplinary efforts to identi fy at-risk individuals as early as possible are underway. Although there have been many studies attempting to predict AD in the preclinical stage (Albert et al., 2003; Jacobson et al., 2002; Johnson et al., 1998; Reiman et al., 1996; Tier ney et al., 1996), there are no well-validated techniques for identifying asymptomatic individuals w ho will eventually develop the disease. Several current techniques can predict the probability of eventual development of AD, but each has significant drawbacks or lim itations. For instance, several studies have demonstrated that genetic risk factor s like possessing 1 or 2 copies of the 4 allele of the Apolipoprotein E gene ( APOE ) may identify those at hi gh risk for developing AD. However, this type of screening may be le ss useful in determining when the disease process begins (Mayeux et al., 2002), and is sometimes positive in people who do not develop AD. Furthermore, approximately onehalf of the patients who go on to develop AD have no copies of the 4 allele (Johnson & Albert, 2003) . Biological markers such as plasma A and tau (Morris, 2005) have also show n some promise in predicting future risk of developing AD, but at this time there are no universally accepted diagnostic markers (Graff-Radford, 2003).


3 There is evidence that structural neur oimaging, particularly measurements of hippocampal atrophy and medial temporal lobe volume, can be helpful in predicting who may develop AD (Jack et al., 1997; Jack et al., 1999; Grundman et al., 2004), but this predictor is more useful in determini ng who will eventually crossover to AD once cognitive symptoms are already present. F unctional imaging techniques support findings suggesting that hypometabolism in a consistent set of brain regions exists in prodromal AD (i.e. hippocampal complex, the anterior a nd posterior cingulate, and the inferior parietal cortex; Johnson & Albert, 2003). Overall, although imag ing techniques have been shown to successfully predict dementia in patients that are already experiencing mild cognitive deficits (Korf et al., 2005), th ese techniques have not proven effective in identifying the beginning, or “preclinical” pha se of the disease. In addition, like genetic testing and screening of biomarkers, neur oimaging procedures are complicated, often invasive, and invariably expe nsive, and therefore may not be well suited for routine screening at the population le vel (Knopman et al., 2001). Because progressive difficulty with memory is considered the earliest sign of AD, neuropsychological evaluation ha s been used as an appro ach to early detection. For instance, standardized memory measures with a delayed reca ll trial have been proven to be effective in detection of memory impairment and crossover to AD (Bondi et al., 1994; Jacobs et al., 1995). Although cognitive te sting is non-invasive and relatively inexpensive, results are complicated by the fact that mild memory impairments are common in the elderly population and so do not always reliably distinguish between pathological and healthy i ndividuals or groups. Anothe r potential limitation of standardized neuropsychological tests of memory in preclinical detection is that we must


4 be cautious when inferring structural or f unctional changes in anatomy on the basis of these tests. Memory tests are only indirect m easures of changes in brain processing; this in part is due to their inherent lack of “process purity.” More specifically, although standardized clinical instrume nts may be sensitive to preclin ical changes in the medial temporal region, performance on these measures is also at least in part dependent upon other cognitive processes (e.g., sustained at tention, working memory, organization, and processing ability) that are mediated by brain structures and systems outside the temporal memory system (e.g., frontal and other cortical structures). Therefore, development of measures that specifically and directly tap into the medial temporal memory system is one promising way of predicting which indi viduals will eventually develop dementia. The Medial Temporal Lobe Memory System (MTMS) Structures in the medial temporal lobe (especially the hippocampus and overlying entorhinal and parahippocampal cortices) play an important role in the acquisition of new memories (Squire & Zola, 1997; Squire & Zola-Morgan, 1991). Research with humans and monkeys has shown that damage to the medial temporal lobe can cause memory impairment without affecting other cognitiv e functions (Squire & Zola, 1997); similar research suggests that the severity of the im pairment increases as a function of extent of damage to the MTMS (Zola-Morgan et al., 1994). Although ma ny cases of human memory impairment point to damage to multiple regions of the medial temporal lobe (Petten, 2004), damage restricted to the hippocampus proper can produce amnesic syndromes, supporting evidence that it is an important anatomic substrate of episodic memory (Cummings et al., 1984; Victor & Agamanolis, 1990; Zola-Morgan & Squire, 1986).


5 The neuropathology (i.e., neurofibrillary tangles and neuritic plaques) of AD typically begins in the tr ansentorhinal area and hippocampal formation, and then progresses to the adjacent medial temporal lobe limbic areas, to ne ocortical association areas, and, finally, to primary sensory and mo tor areas (Braak et al., 1993). Because of the role of the hippocampus and surrounding stru ctures in new learning and the formation of new memories, the early clinical signs of AD include memory impairment, with deficits in language, executive functions, a nd attention becoming ev ident as the disease progresses (Salmon & Bondi, 1999). Normal aging is also associated with cogni tive decline in a number of domains. For example, there are notable age-associat ed deficits in episodic memory, attention, working memory, and spatial learning (Drisc oll et al., 2003). Many studies are already underway investigating the relationship between hippocampal volumes and memory ability in healthy individuals, with vari able results (for a review, see Petten, 2004). Although it appears these imaging techniques may be beneficial for study when clear pathology is present, there still is the need to develop measures that are sensitive to subtle neural damage in at-risk individuals Mild Cognitive Impairment (MCI) This heterogeneous term has evolved to represent a transitional state between normal aging and dementia (Petersen, 2003). Am nestic MCI is a more specific term that describes patients who have memory complaints that exceed those expected during normal aging, but who are not demented according to DSM-IV or NINCDS-ADRDA criteria (Flicker et al., 1991; Petersen et al., 1999). The amnes tic form of MCI is the most commonly studied, appears to be the most degenerative in nature, and it is frequently thought of as a prodromal state of AD (P etersen, 2003). Amnestic MCI individuals


6 typically present with forgetfulness and recen t memory loss, but otherwise have intact cognitive function and perform normally in activ ities of daily living (Welsh et al., 1991). Amnestic MCI is formally defined by the followi ng five specific criteria: (a) Presence of a memory complaint, preferably corroborated by an informant, (b) presence of an objective memory impairment on neuropsychol ogical testing, (c) normal (non-memory) general cognitive function, (d) intact activi ties of daily living, and (e) not demented (Petersen, 2003). The diagnosis of amnestic MC I therefore requires c onsiderable clinical judgment, but has been shown to be distinct from AD and normal aging (Grundman et al., 2004). Studies suggest that individuals with am nestic MCI progress to clinically probable AD at a rate of approximately 10% to 15% per year (Petersen & Morris, 2003). The study of MCI individuals holds much pr omise in examining early predictors of AD, and has been the focus of a great deal of recent research. Patients with amnestic MCI have been shown to have hippocampal volumes that were intermediate between those of controls and patients with AD (Grundma n et al., 2004; Korf et al., 2004). An interpretation of the above data is that the disease ha s not yet spread from the hippocampal formation and surrounding structures into the neocortical association areas. Although it would be ideal to longitudinally follow asympt omatic individuals until the onset of the disease, this task would pose a great challenge in both duration and expense. It seems more feasible to study patients in a phase when pathology has already begun, but before a diagnosis of AD is warranted (Winblad et al., 2004). An important step must be taken before testing clinical tools that propose to detect underlying hippocampal pathology in MCI populations: We must firs t validate these tools in normal aging


7 populations. The present study offers one such tool and takes a first step toward the validation process. Cognitive Mapping Theory and the Role of the Hippocampus in Spatial Mapping Cognitive mapping theory proposes that internal representations (cognitive maps) of external environments are c onstructed so that an organism can effectively interact with and navigate through the worl d (O’Keefe & Nadel, 1978). This theory posits that these mental maps consist of internal repr esentations of geometric relations among environmental stimuli, formed to help the or ganism effectively define its position within the environment (Nadel, 1991). The act of purposefully moving through the environment based on this stored spatial representation ha s been termed spatial navigation (Roche et al., 2005). “Allocentric” navigation is a term that describes viewer-independent navigation and is also known in the liter ature as cognitive map-based navigation or wayfinding. Clearly, the ability to remember important elements of the external world (e.g., food caches, locations of predators, shelte r) and the locations of these elements in relation to each other an d to the organism provide great ad aptive and survival value in the animal kingdom. In humans, such abilities co ntribute heavily to independent function in everyday environments. The task of understandi ng the development of this ability and the corresponding neuroanatomic correlates has been extensively studied, and there exists an abundant amount of literature speculating how organisms form and retrieve cognitive maps of novel environments. A popular view is that the hippocampus is critical for creating and maintaining spatial maps (O’Keefe & Nadel, 1978). Studies investigating food-st oring birds suggest that successful retrieval of their food caches is dependent upon spatial memory (Sherry et al., 1981; Shettleworth, 2003). Evidence from these studies also suggests hippocampal


8 enlargement in caching species of birds when compared to their non-caching counterparts (Krebs et al., 1989; Sherry et al., 1989). Similarly, Maguire an d colleagues (2000) found that the hippocampal volumes of London taxi cab drivers were larger than those of control subjects that did not drive taxis. Moreover, hippo campal volume was correlated with the amount of time spent in the profession, with driver s having longer histories in the profession demonstrating th e largest hippocampal volumes. Numerous studies have also shown that animals with hippocampal damage display spatial navigation impairments (Morris et al., 1982; Sutherla nd et al., 1982). In humans, lesion studies also suggest that the hippo campus is involved in spatial memory (e.g., Astur et al., 2002; Bohbot et al ., 1998; Frakey, 2005). Given th is set of data, atrophy of the medial temporal lobe structures that is seen in aging may explain why clinical observation of older adults who are otherwise cognitively intact, could find it increasingly difficult to navigate (find their way) in unfamiliar environments and to learn about new places. Testing Allocentric Spatial Memory and the Morris Water Maze (MWM) One of the most reliable tests used to te st allocentric (i.e., observer-independent) spatial learning and navigation is the Morris Water Maze (MWM; Morris, 1981), a paradigm that has been mainly used with r odents. In this task, rats are required to navigate to a hidden platform in a pool of mil ky water. The task is structured so that the animal cannot use a specific route to navigate to the platform--it starts at different locations within the pool on each trial. Inst ead, it uses the rela tionship among distal spatial cues to find the platform. On the fina l trial of the task (a “probe” trial), the platform is removed in order to test whethe r the rat has formed a representation of where the platform is located. Healthy young rats learn to find the platform efficiently and


9 accurately. In contrast, rats with hippocampa l damage show severe impairments in the use of a cognitive mapping strategy to find the platform when compared to shamoperated rats (Morris et al, 1982; Sutherland et al., 1982). Fu rther, aged rats perform more poorly on the MWM than do their younger counterparts (Wilson et al., 2003). Comparative Neuropsychology The current study attempts to take a “co mparative neuropsychology” approach to developing more sensitive and specific tests of hippocampal function in humans. That is, this study attempts to further our understa nding of brain-behavi or relationships by adapting experimental paradigms, developed fo r use in animal rese arch, to testing in human clinical populations (Oscar-Berman, 1994). An example of the comparative neuropsychological work is the recent applic ation of spatial navigation tasks adapted from the animal literature to the study of human spatial cognition (Astur et al., 1998; Driscoll et al., 2005; Moffat et al., 2001; Thomas et al ., 2001). Computer-generated environments used in these investigati ons allow for navigation through space without losing experimental control, while preserving features of real-world environments that are lacking in traditional tabletop tests. Virtual environment technology also permits accessibility to more diverse populations who may otherwise be incapable of participation (e.g., non-ambul atory individuals). The research study described in this paper utilizes an adaptation for human use of the Morris Water Maze (MWM), called th e Computer Generated Arena (CG Arena; Jacobs et al., 1997; Thomas et al., 2001). This task mimics the procedural conditions of the MWM and assesses spatial learning, memory , and navigation in humans. Studies of CG Arena properties suggests that data obtaine d from humans in the virtual environment are analogous to those obtained from rats in the MWM. Normal subjects use distal spatial


10 cues to navigate to the target, and re arrangement of these cues produces profound impairment in ability to navigate to the hidden target (Jacobs et al., 1998). Subsequent findings from studies of the CG Arena with clinical populations have also paralleled many of the MWM findings in the various conditions used with rats. Skelton and colleagues (2000) tested the CG Arena with patients who had suffered mild to moderate traumatic brain injury and found that they demonstrated impaired performance when compared to normal matc hed controls. Compared with older adults, other studies have found that younger adults na vigate to the hidden target in less time over trials and spend a greater proportion of time in the co rrect quadrant during the probe trial (Thomas, 1999; Laurance, et al., 2002). Ou r laboratory recently demonstrated that unilateral anterior temporal lobectomy patie nts performed significantly impaired on CG Arena as compared to age-matched contro ls (Frakey, 2005). To date, comparisons among CG Arena performance, real life navigation, and self-reported impairments in spatial abilities have not yet been examined. Furt her, there have been very few studies examining age-related allocentric spatial processing differences. Further Assessment of Spatial Ab ilities in Aging Populations Numerous studies document the existence of age-related changes in human spatial cognition and behavior (for a review, see Kirasic, 2000). These age-related changes include, but are not limited to, the learni ng of novel environmen tal layouts (Kirasic, 1991), the learning of routes (B arrash, 1994), and abilities on mental rotation and spatial visualization tasks (Hertzog & Rypma, 1991) . One complication in studying spatial cognition in humans is that many distin ct constructs (e.g. wayfinding, landmark knowledge, spatial cognition, sense of dir ection, mental rotation, point localization, route-based learning) can fall unde r the general rubric of “spa tial abilities”. Adding to the


11 confusion, several different methods, incl uding psychometric tests, real-world environments, virtual environments, and self-r eport sense of directi on questionnaires, are used to study these different constructs. Two commonly described ways of learni ng the layout of a novel environments are wayfinding (i.e. cognitive mapping or allocentr ic navigation) and e gocentric navigation (i.e. route learning or route following). E gocentric navigation (so called because it is viewer-dependent) is also know n as route following or route learning because it has its foundations in route-based knowledge. In egoc entric navigation, the organism follows a predetermined series of direc tions and turns with the goal of moving toward a specific targeted location. In one study of this navi gational ability, Moffat and colleagues (1998) found significant positive correlations between scores on psychometric tests of mental rotation, map learning, and spatial orientati on, and performance on egocentric virtual maze learning in humans. In contrast, allocentric navigation relies on a view er-independent, external perspective (a map-like or aeria l view) that is thought to allo w direct access to the overall spatial layout (Shelton & Gabriell, 2001). In one study of this navigational ability, Astur and colleagues (2004) found a significant pos itive correlation between mental rotation ability and performance on an allocentri c virtual navigation task. Although both egocentric and allocentric navigation recruit common networks of brain areas, allocentric cognitive mapping is more sensitive to hippocam pal function (for review see Roche et al, 2005). Impairment of egocentric navigational ab ility is typically associated with damage to inferior medial occipital and medial tem poral region structures (Barrash et al., 2000). Deficits in allocentric navigation (as meas ured by VMWM) have been demonstrated in


12 patients with hippocampal damage when compared to age matched controls (Astur et al., 2002). In clinical practice, neuropsychologist s commonly measure “spatial ability”, but rarely engage the patient in navigational tasks of any sort. A large body of data (e.g., Nadolne & Stringer, 2000) suggests that trad itional psychometric tests of spatial ability (e.g. tools that measure visualization, spat ial orientation, mental rotation, and map reading) tap into domains of visuospatial skills that are not necessarily crucial to spatial navigation ability (i.e., the ability to m ove around a new environment and learn its layout). For instance, Sonnenf eld (1985) demonstrated th at paper and pencil spatial performance was independent of true wayfindi ng ability. In this st udy, participants from Southeast Alaska, including professional guide s, fishing boat captains and pilots, were found to have some of the poorest performan ces on a battery of paper and pencil spatial tests. Additionally, Kirasic (1991) also demonstrated no correlation between cognitive task performance in a laboratory and perf ormance in a real-world novel environment. Similarly, Kirasic (1988) found no significan t relationship between performance on traditional paperand pencil-based psychometric measures and navigation ability in elderly individuals. Overall, then, the ability to navigate through the environment appears to be an ability independent of those abiliti es measured by traditional psychometric tests, and direct inferences cannot be made about one from the other (Maguire et al., 1999). Because traditional paper-and-pencil instruments have not consistently been proven to be sensitive to underlying impairment, inve stigations have also evaluated real-world environmental abilities. For clinical purposes, however, it is not feas ible to subject all patients who present with navigation and topographic difficulties to a large-scale


13 environmental task. Common criticism of expe rimental measures is that they lack ecological validity.1 Here, we define ecological validity in a clinical setting as the extent to which performance on tests can be generali zed to the real worl d or an individual’s everyday environment (Nadolne and Stringer, 2000). There have been relatively few studies investigating the ecologi cal validity of our current m easures of spatial skill and even fewer validating virtual navigation with navigation in a real wo rld environment. As discussed above, our current psychometric measur es seem to lack the ability to generalize performance in a natural setting requiring spatial skill. Another approach to the assessment of sp atial abilities has been to use self-report measures of spatial competence. Some self -report measures of environmental spatial abilities (termed “sense of direction” in one segment of this literature) show high positive correlations with other measures of the same construct (Bry ant, 1982). For instance, Kato and Takeuchi (2003) demonstrat ed that self-reported good sens e of direction is correlated with better performance on a real-world ro ute-learning task. More recently, Hegarty and colleagues (2002) found that alt hough self-reported sense of di rection is correlated with map drawing, real-world landmark knowledge , and video and virtual egocentric navigation, it is not significan tly correlated with pencil and paper psychometric tests. This weak correlation between self-repor t and psychometric m easures has been demonstrated in other studies as well (e.g., Sholl, 1988; Bryant, 1982). No studies have examined the relationship between self-reporte d sense of direction and virtual allocentric navigation. In this st udy, we developed a sta ndardized self-report 1 The term ecological validity has a variety of meanings. Within the research domain, it is the extent to which findings can be generalized to the "real world" ; the ability to generalize from a test to an everyday criterion (Nadolne and Stringer, 2000); or the extent to which performance in the laboratory resembles that of real-life everyday behaviors.


14 scale of changes in spatial abilities, the Spatial Activities in Environment Questionnaire (SpAcE-Q, please refer to the de scription in the methods sec tion), in order to detect a more clinical impairment in spatial navigation abilities. In summary, our review of the literature suggests that although there are currently many ways to assess spatial ability, the relationships am ong these methods are largely unknown. The use of virtual environments has made it possible to study wayfinding performance in a controlled setting, and the CG Arena task appears to measure abilities that are not measured by standa rd clinical psychometric in struments. The study described here will help lay the foundation for futu re study of spatial navigation in clinical populations by helping to devel op tools that are sensitive to underlying medial temporal lobe pathology. The eventual goal is to apply measures validated in this study to clinical identification of patients at risk for developing AD. This study aimed to evaluate the rela tionships among psychometrically defined spatial cognition abilities, self-report sp atial navigation impairment, real-world environmental knowledge, and virtual spatial navigation ability. The three major aims of the study were to empirically determine (1) whether there are age-group differences in spatial navigation performance, (2) the re lationship between a measure of spatial navigation and our current clinical measures of spatial navigati on and cognition, and (3) the relationship with navigation in real space wi th that in virtual space. To examine these aims, healthy young adults and healthy old adults were examined. Specific Aim 1 To characterize differences in spatial navigation performance between two groups (healthy young and healthy elderly). Based on the extant literatu re, it was hypothesized


15 that elderly participants would show impa ired performance on the C-G Arena compared to healthy young participants. (healthy old pa rticipants < healthy young participants). Specific Aim 2 The second specific aim was to examine the relationship between performance on clinical measures of spatial cognition (using psychometric measures of spatial cognition and clinical assessments of spatial naviga tion impairment) and performance on a virtual spatial navigation task. Consistent with the premise of convergent validity (i.e. correspondence between similar c onstructs) and discriminant validity (i.e. differentiation of dissimilar constructs) we pr edicted relationships using th ree methods as an empirical guide. First, available litera ture indicates that performan ce on a questionnai re assessing symptoms of real world spatial navigation w ould be correlated with performance on the C-G Arena. Although some studies suggest co rrelations among traditional psychometric measures of spatial ability and performance in computer generated environments, other studies indicate no such corre lation, and suggest that pape r-and-pencil tests are not sensitive to underlying spatial navigation impair ment. Overall, the literature suggests that mental rotation will be correlated with the CG Arena. Second, based on the anatomy of the hippocampus and the involvement in spa tial mapping and navigation, we predicted that performance on all of our psychometric measures of spatial cognition would not be correlated with performance on the C-G Arena. Based on a more liberal view of the role of the hippocampus, we predicte d that all measures of non-verbal memory would share some overlapping variance with the CG Aren a, and thus would correlate with Arena performance. Third, based on a cognitive com ponents construct of spatial ability, we predicted that mental rotation, US Map Test , Florida Map Test, and the NAB Map test would share overlapping variance tapped by the CG Arena, and thus would correlate with


16 Arena performance variables. This last predic tion is based on the assumption that all of these measures load on a common ability to mentally store and manipulate images in order to solve a problem. Specific Aim 3: To examine the relationship between navigation in C-G Arena space with navigation in real space. We predicted that better performance on a real world navigation task would be associated with better performance on the C-G Arena.


17 CHAPTER 2 METHODS Participants Participants were twenty-four indivi duals (12 age 18-23, and 12 age 65-84) recruited from Gainesville, Florida and su rrounding areas. Recruitment for the old group took place via community advertisements (e .g. newspaper, flyers, and newsletters). Recruitment strategies for young healthy adults focused mainly on University of Florida advertisements (e.g. flyers). The main goal of recruitment was to obtain a heterogeneous and representative sample of commun ity-dwelling old adults, without cognitive impairments, and a sample of young adults matched on the basis of sex, education, and ethnicity. All potential participants were screen ed by telephone to determine whether individuals met the following exclusionary criteria: (1) dementing illness or other neurological disease, (2) hist ory of significant head injury, (3) major medical illness, (4) severe uncorrected vision or hearing impair ments, (5) history of inpatient psychiatric treatment, (5) history of clinically signifi cant drug or alcohol abuse, (6) unwillingness to participate in two testing sessi ons, or (6) inability or unwillingness to walk 1/3 of a mile on three separate times within an hour tim e period, or unwillingness to participate by being escorted in a wheelchair. As a result of screening, all partic ipants in both groups were cognitively intact a nd had no history of neurological insult or psychopathology. Participants gave written informed cons ent according to university and federal regulations. All participan ts who completed the res earch protocol received $50.


18 A total of twenty-nine participants we re initially enrolled in the study and completed all study procedures. The data from one participant were not included in the final analysis due to a depression score in the clinical range . Additionally, four participants were excluded from the final anal ysis because they did not meet criteria for being cognitively intact (as de scribed in the statistical anal ysis section below). Finally, one participant was not included in the third se t of statistical analyses described in the Results section because she was not able to complete the delayed recall portion of the Florida Route Learning Task. Demographic variables for the remaining 24 participants (7 females and 5 males in the old group and 6 males and 6 females in the young group) are shown in Table 2.1. The racial composition of the sample was 9 Cau casian and 3 Hispanic males and females for the young group and 100% Caucasian for the ol d group. The two groups did not differ in education [ t (22) = .306, p > .05] or IQ [ t (22) = .016, p > .05]. Table 2-1 Demographic Characteristics of Experimental Participants by Group Measure Young Adults Mean (SD) Old Adults Mean (SD) Number of Participants N = 12 N = 12 Age 20.83 (1.47) 72.08 (6.87) Education 14.00 (1.28) 14.25 (2.52) IQ 116.67 (9.40) 116.75 (15.42) Note: IQ= Intelligence quotient as measur ed by the Wechsler Abbreviated Scale of Intelligence (WASI) Procedures Following the completion of the abovementioned telephone screening and consent procedures, all eligible participants were invited into the psychology laboratory for the assessment procedures. Testing was conducte d over two sessions. During the first session, we conducted an initial neuropsychologi cal assessment (battery described below) with the aim of screening for dementing illness.


19 First testing session. At this session, the participan tÂ’s overall cognitive functioning was assessed using the Neuropsychological Sc reening and Assessment Battery described below. Additionally, participan ts completed mood inventories and provided lists of their current medications. Second testing session. Within one month of the first testing session,1 participants were administered the spatia l cognition psychometric battery described below, as well as self-report measures of spatial navigation impairment, a computer-generated navigation task, and a real-world navi gation task. All testing was conducted by one examiner. Neuropsychological Screening and Assessment Battery The Neuropsychological Screening and Assessment Battery consisted of standardized tests of neur ocognitive functions. The elem ents of the battery are summarized below (Table 2-2). All tests ar e from peer-reviewed sources, and each measures a particular neuropsychological dom ain or function. These measures represent some of the most commonly used neuropsychol ogical instruments in the domains likely to indicate early loss of cogni tive functioning in dementia, and in domains likely to remain intact in early AD. With an eye towa rd extending this project to old adults with mild cognitive impairment (MCI), the compos ition of this test battery was determined by a desire to distinguish healthy old adults fr om old adults with MCI, paying particular attention to verbal and nonverbal memory functioning and subjective memory complaints that are fundamental to the MCI diagnosis. The domains chosen are the same as those reflected in the Consortium to Establish a Registry for AlzheimerÂ’s Disease (CERAD) battery (Morris et al., 1989). 1 the exact time interval between test sessions varied depending on each participantsÂ’ schedule


20 Table 2-2 Neuropsychological and Clinical Screening Measures Overall ConstructMeasureDescri ption of Test MeasureSource General Mental Status Mini Mental State Exam (MMSE) Estimation of dementia severit y Folstein et al., 1975 Observer/ Interview Ratings Clinical Dementia Rating Scale (CDR) Disease severity rating: informant & patient subjective, objective elements Morris, 1993 Functional Assessment Memory Assessment Centers-Questionnaire ( MAC-Q ) Subjective memory complaints Crook et al., 1992 Intellectual Functioning Wechsler Abbreviated Scales of Intelligence (WASI) Vocabulary, Similiarites, Block Design, Matrix Reasoning Predicts general intelligence, verbal and performance abilities Wechsler, 1999 Hopkins Verbal-Learning Test-Revised (HVLT) Verbal list memory test that assesses learning, immediate and delayed recall, and reco g nition. Brandt & Benedict, 1971 Wechsler Memory ScaleIII -Logical Memory I, II Measure of verbal rote and memor y Wechsler, 1997 Brief Visual-Spatial Memory Test-Revised (BVMT-R) Measure of visual memoryBenedict, 1998 Rey-Osterrieth Figure Test – Immediate and Delay Figural Memory; Measure of Visuoconstructive ability, incorporates immediate & dela y recall. Osterrieth, 1944 Boston Naming Test-2nd Edition (BNT) Confrontation naming using large ink drawings Goodglass & Kaplan, 2001 Controlled Oral Word Association (COWA) Verbal fluency to alphabet letter (FAS). Spreen & Benton, 1977 Category Fluency Animals WAIS-III Digit SpanAttention spanWechsler, 1997 Trail Making Test A & B Visuomotor speed, tracking and attention Benton & Hamsher, 1976 Geriatric Depression Scale (GDS) Self evaluation assessing elements of depression Yesavage et al., 1983 Beck Depression Inventory -2nd Edition (BDI-II) Self evaluation assessing elements of depression Beck, Steer & Brown, 1996 Attention/ Concentration/ Working Memory Mood Tombaugh et al., 1999 Memory Functioning Language Functioning Verbal fluency to a semantic category in four 15” quarters.


21 Spatial Cognition Psychometric Battery The elements of this battery are describe d below (Table 2-3). The rationale for the inclusion of the chosen measures of spatial c ognition is based on two principles. First, the particular domains of spatial cognition tested are derived fr om prior work establishing them as theoretically and statistically discrete categories of spatial ability (Dabbs et al., 1998; Kirasic, 2000; Voyer et al., 1995). S econd, the particular spatial cognition assessment instruments represent either expe rimental measures specially developed to evaluate the domains of interest, or well-est ablished and widely used measures of those domains. Table 2-3 Spatial Cognition Psychometric Battery Cognitive DomainMeasureSource Visuospatial FunctionJudgment of Line Orientation (JLO) Benton, Sivan, Hamsher, Varney, & S p reen ( 1994 ) Visuomotor FunctionRight-Left Orientation TestBenton (1959) WASI Block DesignWechsler, 1999 Clock DrawingFreedman, et al., 1994 Spatial Cognition – Mental Rotation Mental Rotation TestVandenberg & Kuse (1978) Spatial Cognition – Visual Memory configuration, location and orientation of figural stimuli Building Memory TestEkstrom (1976) US Map Test Florida Map Test Spatial Activity in Environment Questionnaire (SpAcE-Q) King, Thomas, & Bauer (2005) Neuropsychological Assessment Battery – Spatial Module: Map Test (NAB) Visuoperceptual/ Visuoconstruction 7/24 TestBarbizet & Cany (1968) Spatial Cognition – Object Location Memor y Spatial Cognition – Geographic Knowledge Santa Barbara Sense of Direction Scale (SBSOD) Hegarty, Richardson, Montello, Lovelace, & S ubbiah (2002) PAR, Inc. (2003) Spatial Cognition – Local Navigation Strategy Spatial Cognition – SelfReported Skills; Clinical spatial navigation impairment


22 Computer-Generated Arena The Computer-Generated Arena (CG Ar ena; Jacobs et al., 1997, 1998) is a computer-based analogue of the Morris Water Maze task (MWM) and is administered on a desktop or laptop computer. The participant is asked to use a joystick to navigate through a virtual MWM in order to find a hidden target. The stimulus environment consists of a circular arena wall located with in a small square “room.” This virtual room is analogous to the circular tank placed within a square room used in Morris’ original experiments. Each of the room’s four walls contains a unique item, such as a picture, a door, a window, or a pattern that together serve as distal cues to assi st the participant in navigation (Figure 2-1). The placement of the wa lls relative to the target remains constant over all the experimental trials. The target is a small square locate d on the floor of the Arena. The task itself is modeled after the classic MWM paradigm. On each trial, the participant starts from a different point in the circular arena. The current CG Arena protocol began with a set of practice trials. During those trials, the target was visible, and the particip ant was asked to use th e joystick to navigate to it as quickly as possible. The target wa s in a different place in the room on each practice trial. Over the course of the practice trials, part icipants were exposed to a minimum of five minutes practice time in order to familiarize themselves with the use of the joystick. Participants were administered the appropriate number of practice trials until the examiner judged that all pa rticipants were starting the ex perimental trials with an equivalent level of understanding of the task, as well as familiarity with the joystick. Immediately after completing the set of practice trials, participants were administered a set of acquisition trials. During those trials, the participants entered into a new virtual room for eight trials. The target was invisible, but remained in the same


23 location across acquisition trials . During the first trial, the participant navigated through the environment until the invisible target was found. Once this occurred, the target became visible (Figure 2-2) and was paired w ith an auditory clicking. The target also “trapped” the participant and made it impossible to move off by use of the joystick, forcing the participant to look around the Arena environment. This procedure was repeated for the remaining seven trials. In the event that the participant was unable to independently locate the target within 120 seco nds, the examiner assi sted the participant. Such assistance was given only on the first two acquisition trials. The starting position within the Arena was randomized for the first six acquisition trials. In order to measure learning in the data analysis, Trial 7 had the same starting position as Trial 2, and Trial 8 had th e same starting position as Trial 1. Immediately following the 8 acquisition tria ls, the participant was administered a probe trial. On this trial, the target was, removed fr om the virtual room, unbeknownst to the participant. This final trial is an an alogue of the standard probe trial in MWM research, in which the animal, knowing where the target “should be ,” repeatedly swims around the anticipated target location, sear ching for it. Upon completion of the probe trial, the participant was presented with a blank screen, indicati ng the end of the CG Arena portion of the testing session. Particip ants were then immedi ately debriefed about the removal of the target on the probe trial. For each trial, several dependent measures , including the length of the navigation path, the latency to find the hidden target, th e time spent in each quadrant of the arena, and whether the target was actually found, were automatically recorded by the CG Arena software.


24 After the probe trial, participants were administered two paper-and-pencil measures related to the CG Arena task: an arena reconstitution task (ART) and an object recognition task (ORT). The ART required participants to r econstitute the CG experimental room by appropria tely placing icons representi ng the four walls of the room, the objects on those walls, and the target onto a sheet of pa per. The ORT required participants to correctly identify and disti nguish objects that were on the walls in the experimental room from a group of objects, some of which were in the room and some of which were not. The entire CG Arena protoc ol, including the ART and ORT tasks, took approximately 30-45 minutes to complete. Figure 2-1. Representations of the C-G experi mental room as seen by the participants Figure 2-2. Representation of target once it is successfully acquired


25 Florida Route Learning Task The Florida Route Learning Task (FRLT) mi micked key procedural conditions of a task developed by Barrash and colleagues (Barrash, 1994; Barrash, Tranel, & Damasio, 1993) originally designed to evaluate patients with topogr aphic disorientation. The FRLT assessed learning of a one-third mile route, rich with visual cues, that wound through the main corridors, clinic areas, and back hallw ays of the ground floor of Shands Teaching Hospital, Gainesville, Florida (Figure 2-3). At the starting point, participants were told that they would be asked to learn a route through the hospital and back to the starting point. They were further inst ructed to pay careful attent ion because immediately upon return to the starting point th ey would have to lead the way through the same route from memory. Participants then followed the ex aminer, who walked through the route at a leisurely pace (approximately 10 minutes to complete the ro ute). Although this pace was rather slow for the young partic ipants, it allowed for equal ex posure time for each group. After the acquisition trial, subjects were inst ructed to lead the way through the route for one test trial. The examiner followed behi nd on the test trial, correcting each error immediately by pointing in the correct direction while stati ng that “the route goes this way.” A delayed recall trial, identical in pro cedure to the test trial, followed 30 minutes after completion of the test trial. Participan ts were not informed that there would be a delayed recall trial. The FRLT consists of 29 intersections with a varied number of decision points (i.e. right, left, and straight turns) . Twenty-two of the intersec tions had two decision points, six had three decision points, and one ha d one decision point. To account for these differences each intersection was weighted based on the number of possible decision points (i.e. 1, 2, or 3). The immediate r ecall FRLT score and the delayed recall route


26 score were calculated by summing up the total number of correct weighted intersection decisions. 14 15 16 17 13 18 12 19 20 21 22 23 24 25 26 27 28 29 END 2 1 3 4 5 6 7 8 9 10 11 START 14 15 16 17 13 18 12 19 20 21 22 23 24 25 26 27 28 29 END 2 1 3 4 5 6 7 8 9 10 11 START Figure 2-3. Schematic representation of th e Florida Route Learning Task (FRLT) Self-Report Measures of Environmental Spatial Ability The Santa Barbara Sense of Direction Scale (SBSOD) is a 27-item self-report measure of environmental spatial ability or “sense of direct ion.” The scale has demonstrated good psychometric pr operties (Hegarty et al., 2001). The Spatial Activities in Environment Questionnaire (SPAcE-Q; King, Thomas, & Bauer, 2005) is a 9-item in-house self-report m easure of clinical sp atial ability. It was used to evaluate changes in spatial abilities over the past six months and to detect the possibility of clinical impairment in spatial navigation abilities. Statistical Analyses Two path-length learning de pendent variables were created by using the formula shown below. The starting point in Trial 1 wa s the same as the starting point in Trial 8. The starting point on Trial 2 was the same as the starting point on Trial 7. This


27 calculation therefore represented a measure of learning. The mean of the two path length learning dependent variables was used in the CG Arena composite described below. DVP_L_learning_1: Path Length Trial 1 – Path Length Trial 8 DVP_L_learning_2: Path Length Trial 2 – Path Length Trial 7 Another path length dependent variable was created based on the following formula, which represents the average devia tion from the optimal path length over each trial. Optimal path length represents the shor test possible distance from the starting point to the target on each trial. DVpath_length: Deviation from optimal DVpath_length: AVG [deviation from optimal path length trials 1-8] Deviation from optimal = [optimal path length – actual pa th length / optimal path length] The sum of target acquisitions was recorded by counting the number of times the participant found the invisibl e target over the course of the acquisition trials.2 Dwell time on the probe trial was created by calculating the propor tion of time the participant spent in the target quadrant (the quadrant where the invisible target had previously been located) during the probe trial. The ORT score was created by summing up the total number of correctly identifie d objects and correctly discerning objects that were not 2 Eleven of the twenty-four participants were inadvertently run under a separate time protocol for the acquisition and probe trials. Eleven pa rticipants were run under Protocol A and thirteen participants were run under protocol B. In Protocol A, participants were allotted 120 seconds on the first two acquisition trials, 80 seconds on the subsequent acquisition trials, and 60 seconds on the probe trial. In Protocol B, participants were allotted 120 seconds all acquisition tr ials and 80 seconds on the probe trial. The data reduction procedures corrected for participants run under the different protocols and allow for direct comparisons of all dependent variables except total numb er of acquisitions. In orde r to correct for this, all targets acquired after 80 seconds on protocol B were not scored (resulting in correction of one score for 2 participants).


28 located in the room. The ART score was created by using a templa te that awards points to objects placed correctly in the room relative to the actual location. Three separate analyses corresponding to a priori predictions were conducted and a significance level set to alpha = .05 was used. Participants we re judged to be cognitively intact if they met the following criteria: (a) had no memory complaints as assessed by the MAC-Q, and corroborated by an informant on the CDR, (b) performed normally for age on all subtests of memory, and (c) scored at or above 26 on the MMSE. Participants categorized as cognitively impaired were excl uded from the data analyses described in this paper. In order to characterize differences in spatial navigation performance between the young and old group a t -Test was conducted to detect a si gnificant main effect of group membership. In order to weigh the various CG Arena dependent variables equally, a composite dependent variable (DV) was created from the mean of the z scores earned on each variable ( path length; path length learning; su m of target acquisitions across invisible trials; dwell time on pr obe trial; ORT score; ART score ). The basic statistical procedures we used to create a composite DV are outlined in Rosenthal (1991). To examine the relationships among CG Arena performance, psychometric measures of spatial cognition, and clinical selfreport as sessments of spatial navigation impairment, bivariate correlations were conducted on the combined group and on each age group separately. Results from the two self-report me asures were summed to create an overall self-reported spatial cognition score. Each of the spatial cognition measures was examined separately.


29 To compare navigation in CG space with navigation in real space, a bivariate correlation between the CG Arena composite and the score on the FRLT. This was conducted on each group separately as well as the combined group.


30 CHAPTER 3 RESULTS CG Arena Performance The first part of our statistical analysis examined differences in overall CG Arena performance between old and young adults. Both the young adult and old adult groups found the target consistently wh en the target was visible on the practice tria ls. Only one old participant reported joystic k experience. To exclude th e possibility that the group differences observed in CG Arena performa nce was secondary to greater joystick experience in the young group, a mean path leng th score was calculated for all visible practice trials. This variable did not reveal reliable age group differences [ t (22) = 1.965, p = .062]. Joystick experience failed to meet the assumptions of ANCOVA and therefore was not used as a covariate in further analyses. A univariate analysis of variance (A NOVA) was performed with individual variables of CG Arena performance as th e independent measures and age as the dependent variable. Three of the CG Aren a variables revealed the expected young group advantage. The young group found the target in shorter distance (path length) [ F (1,22) = 39.61, p < .001] (Figure 3-1), more often (numbe r of target acquisitions, Figure 3-2) [ F (1,22) = 67.71, p < .001], and spent a greater percentage of time in proximity to the target on the probe trial [ F (1,22) = 17.28, p < .001] (Figure 3-3). The old and young groups did not differ in their performances on the Object Recognition Task (ORT) [ F (1,22) = 2.19, p = .153], path length learning variable [ F (1,22) = .108, p = .745], or on the Arena Reconstruction Task (ART) [ F (1,22) = 3.17, p = .089]. However, the old group


31 placed the target in the correct quadrant on th e ART significantly less often than did the young group [ t (22) = -2.275, p = .033, d = 0.93]. A 2 (Group) x 8 (Trials) repeated measures ANOVA conducted on the path leng th over the eight trials detected no significant group x trials interaction [ F (1,7) = 1.08, p = 3.74] 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 12345678TrialMean Path Length Ratio Deviation Old Young Figure 3-1 Mean SE (± 1) path length deviation ratio of young and old participants over the eight acquisition trials. Compared to old adults, young adults were able to locate the target in a sign ificantly shorter distance. Young Old 8 7 6 5 4 Total Target Crossings Figure 3-2 Mean SE (± 1) target crossings of young and old participants. Compared to old adults, young adults were able to loca te the target significantly more often.


32 Young Old 100 80 60 40 20 Mean % Time Spent in NW Quadrant Figure 3-3 Mean SE (± 1) percentage of time spent in the northwest quadrant on the probe trial. Compared to old adults, young adults spent more time in proximity to the target on the probe trial. The composite CG Arena variable, whic h represented overall performance, was subjected to an independent samples t -test to test for group di fferences in allocentric navigation. Significant group differences were found [ t (22) = -5.142, p < .001, d = 2.19]. Overall, elderly participants performed significantly worse on the Arena composite than did their younger counter parts (Figure 3-4). Young Old 0.75 0.50 0.25 0.00 -0.25 -0.50 -0.75 CG Arena Composite Score Figure 3-4 Mean SE (± 1) CG Arena composite performance by group [ t (22) = -5.142, p < .001, d = 2.19].


33 Correlations Between CG A rena and Clinical Measures The second part of our stat istical analysis examined the relationship between standard paper-and-pencil measures of spatia l ability and the CG Ar ena. Table 3-1 shows the mean scores obtained by old and young adul ts on each of the spa tial cognition tests. Table 3.2 presents the correlation coefficients for the psychometric tests, spatial ability self-report questionnaires, and the CG Arena composite variable. In the old adult group, the Florida Map Test was signi ficantly correlated with overall CG Arena performance. In the young adult group, performance on the Ment al Rotation task (r = .613, p < .05) and performance on the US Map test were sign ificantly correlated with overall CG Arena performance (r = -.648, p < .05). Table 3-1 Mean Performance (SD) on th e Cognitive Tests as a Function of Age Mean (SD)Mean (SD) Young AdultsOlder Adults t d Mental Rotation 20.00 (9.80)8.58 (5.74)-3.48**1.42 Florida Ma p Vector#1.51 (1.11)0.95 (.30)-1.680.69 US Ma p Vector#1.24 (.42)0.88(0.28)-2.46*1.01 RL Orientation 19.58 (0.79)19.00 (2.37)-0.8070.33 JOLO 25.50 (3.42)22.50 (3.50)-2.12*0.87 REYO Immediate 23.17 (4.96)16.70 (6.30)-2.79*1.14 REYO Delay 22.17 (5.50)15.50 (6.12)-2.80*1.15 Block Design 53.25 (14.31)33.08 (12.01)-3.74**1.53 NAB Map 9.08 (1.78)9.36 (1.43)0.4130.17 BVMT Delay 11.41 (.79)8.08 (1.93)-5.54**2.23 7/24 Delay 6.41 (.67)5.66 (1.83)-1.340.54#Note. US and Florida Map Vectors represent average deviation from correct location * Significant at p<.05 level. **Significant at the p<.01 level.


34 Table 3-2 Correlations between CG Aren a Composite and Clinical Measures Subjects Self Report Bldg Mem MR FL Map US Map RL Orien JLO REYO Immed REYO Delay BD NAB Map BVMT Del 7/24 Del Old-.*.349.216-.344.14.091-.273-.20.138-.269 Young .394.329.613 *-.516-.648*-.23.443-.077-.252.44.422.173.305 Note. Correlation significantly different from r = .00, at **p < .01(2-tailed); *p < .05 (2-tailed) Relationship Between Navigation in CG Arena Space and Real Space The third part of our statistical analysis examined the relationship between navigation in real space and that in CG Arena space. Table 3-3 presents correlation coefficients between the Florida Route Learning Task (FRLT) and the CG Arena composite variables. As can be seen, the only significant correlation was between the FRLT delay ( r = .592, p < .05) and the CG Arena composite in the young group. Table 3-3. Correlations between CG Arena and FRLT SubjectsFLRT ImmediateFLRT Delay Old.11.095 Young.331.592* A separate analysis was performed comp aring young and old adults on the real world navigation task. Significant group di fferences were found between the young and elderly group on the FRLT immediate recal l [t(21) = -3.172, p = .001, d = 1.53] and the FRLT delayed recall [t(21) = -2.056, p = .021, d = 1.17]. Overall, elde rly participants performed significantly worse on this task than did their younger counterparts.


35 Young Old 62.00 60.00 58.00 56.00 54.00 52.00 Mean FRLT Score FRLT Delay Total FRLT Immediate Total Figure 3-5 Mean Florida Route Learning Ta sk performance by group. Immediate recall [ t (21) = -3.172, p = .001, d = 1.53] and Delayed recall [ t (21) = -2.056, p = .021, d = 1.17]. Sex Differences Although not the focus of this current study, strong sex differences have been found in many previous studies of spatial cognition and naviga tion (see, e.g., Astur et al., 1998; Sandstrom et al., 1998). We thus c onducted statistical an alyses to explore differences between males and females on our tasks. No sex differences were found on the practice visible path length variable that was used as a measure of joystick literacy [ t (22) = -.009, p > .05], or on the overall CG Arena composite [ t (22) = 1.075, p > .05]. Furthermore, males and females did not di ffer in their performance on any of the cognitive tests of spatial ability. Table 3-4 shows the mean scores obtained by males and females on each of the cognitive tests.


36 Table 3-4 Mean Performance (SD) on the Cognitive Tests as a Function of Gender Mean (SD)Mean (SD) FemalesMales t d Mental Rotation 10.77(8.38)18.45(10.04)2.040.83 Florida Ma p Vecto r #1.29(1.07).53(1.07)-0.340.71 US Ma p Vecto r #1.09(.46)1.02(.32)-0.480.18 RL Orientation 19.15(2.23)19.45 (1.04)0.410.17 JOLO 23.16(3.48)25.00(3.90)1.230.50 REYO Immediate 20.04(7.99)19.82(4.37)-0.0820.03 REYO Delay 19.42(8.25)18.14(4.32)-0.470.19 Block Design 44.54(17.78)41.55(15.59)-0.430.18 NAB Map 9.17(1.95)9.27(1.19)0.160.06 BVMT Delay 9.46(2.72)10.09(1.51)0.680.29 7/24 Delay 6.08(1.26)6.00(1.61)0.160.06#Note. US and Florida Map Vectors represent average deviation from correct location * Significant at p<.05 level. **Significant at the p<.01 level.


37 CHAPTER 4 DISCUSSION This study sought to evaluate various form s of spatial cognition and navigation in young and old adults and to examine their relati onship to real world measures of spatial navigation impairment. We used the CG Aren a, a validated computer analogue of the MWM, to evaluate age-related differences in place learning and allocentric navigation. To assess spatial mapping overall, we deve loped a composite score by factoring in several variables of CG Arena performance. Consistent with previous findings, results from this study clearly demonstr ated that overall, old adults do not navigate as effectively as young adults in CG space. These differences were not due to problems in mastering the procedural demands of the task, but appe ared to be due to differences in place learning and memory (Skelton et al., 2000). Performance on all visible practice trials was comparable, all participants reported they understood the instructions, only one old participant reported familiarity with the joysti ck, and all participants were trained until they mastered the use of the joystick. We also examined CG Arena variables independently to assess spatial mapping similar to established techniques in the anim al and human literature (Morris et al, 1982; Skelton et al., 2000; Thomas et al., 1999; Wils on et al., 2003). This data demonstrated that young adults found the target more often a nd with shorter traveling distance than did old adults. Additionally, on th e probe trial, young adults spent a greater proportion of time searching the target quadran t than old adults. This patter n of results indicated that


38 young adults learned, remembered, and navigated to the invisible target more effectively than old adults (Laurence et al., 2002). The data failed to demonstrate group di fferences in the object recognition task (ORT), an independent paper-and-pencil recognition task admi nistered after the completion of the CG Arena. Although the ORT measures recognition of objects that were located in the room, it does not measure reconstruction of the spatial relationships among the objects, and thus likely taps differe nt memory representations than those used for spatial navigation. The young and old groups did not significantly differ in the distance traveled (path length) learning variab le. Since trial 1 and trial 8 have the same starting point, as well as trial 2 and trial 7, path length differences between trials were obtained. Although the young group demonstrated more efficient navigation as evident by shorter path length over trials, these resu lts indicate that the ra te of learning between groups did not differ significantly. This result is not surprising in light of the fact that both groups were healthy adults. A depressed le arning curve would be predicted if tested in a patient population in which the neurol ogical substrate for learning was damaged (e.g., in patients with MCI or early AD; Pe tersen, 2000.). This resu lt is also partly consistent with performance on visual me mory tests in this study. Old and young groups did not differ significantly on the rate of learning on the Brief Visual Memory Test (BVMT) [ t (22) = 1.47, p > .05], though young adults did perform significantly better on the delayed recall trial [ t (22) = -5.4, p < .001]. Analysis of the arena reconstitution task (ART), which requires participants to reconstitute the CG Arena experimental room from memory, did not reveal significant group differences. However, the old group placed the target in the correct quadrant on the


39 ART significantly less often than the young gr oup, further supporting th e idea that they did not have an accurate memory representati on for the location of the invisible target. Taken collectively, these data suggest that wh ile old adults are able to (a) reconstruct cognitive maps of their environments as well as young adults, (b) recognize objects that were in the environment as well as young adults, and (c) demonstrate a comparable learning curve, they are not as effectiv ely able to navigate in a novel spatial environment. That is, young adults are better able to navigate (f ind their way) using spatial maps of the environment, and are ab le to remember a particular place in that environment more accurately than old adults. These results are compatible with previous findings (Laurance et al., 2002; Thomas et al, 1999) in that old adults appear to acquire an accurate reconstructive map of the CG Arena environment as well as young adults, but do not learn, remember, and navigate to the i nvisible target as eff ectively as young adults. Examination of the relationship between pape r-and-pencil measures of visuospatial ability and CG Arena Performance revealed that, while three psychometric paper-andpencil measures were significantly correlated w ith CG Arena performance, a majority of the measures were not. The significant corr elations that were found were of modest proportions; performance on paper-and-pencil measures of spatial cognition accounted for only 25% 40% of the variance in CG Aren a performance. This finding suggests that, while some tests may share some overlappi ng spatial ability, CG Arena appears to measure a unique skill that is not otherw ise accounted for by our current clinical measures. The results of this study demonstrate that performance on the mental rotation task correlates positively with overall CG Arena performance, which suggests that the two


40 tests share overlapping information processing demands. This result is consistent with past research that has demonstrated signi ficant correlations with the probe trial on a virtual MWM task and mental rotation (Ast ur et al., 2004). The observed correlations tended to differ and vary between age groups. Further work is needed to more clearly elucidate the extent to which CG Arena ta ps basic spatial cognition skills, including mental rotation, visual object memory, and complex visuoperceptual skill. In future studies, it would be useful to include CG Arena and FRLT performance in a factor-analytic study of spatial ability th at includes both paper-and-pencil and selfreport measures of spatial c ognition. While performance on our paper-and-pencil tests may contribute to varying degrees to the same spatial/navigational factor that is critical for CG Arena, this could not be examined in this present study due to the number of participants. In the past, factor analytical studies have described multiple spatial factors (Allen et al., 1996). Two of the tests assess visual memory or the ability to remember the configuration, location and orient ation of figural stimuli (Ekstr om et al., 1976), and all of the tests may share a distinct spatial factor that would emerge in future analyses. However, because the correlations between CG Arena and current measures were scarce and varied, it appears that while some pape r and pencil measures capture abilities that may be used in spatial navigation, there is a large amount of variance that is unaccounted for. For this reason, direct assessment of spa tial navigation may prove to be a useful tool in the clinical evaluation of diso rders of spatial cognition and memory. Interestingly, participantsÂ’ self report of their sense of direction and navigational ability did not correlate with CG Arena perf ormance. This finding suggests that people may not be able to accurately predict their actua l spatial navigation abil ities. Past research


41 has demonstrated that self-reported spatial ab ilities (termed sense of direction) were unrelated to paper-and-pencil psychometric tests, knowledge acquired from maps, and virtual egocentric navigation environments (Hegarty et al., 2002). Our findings indicate that self-reported sense of direction measures are unrelated to perf ormance in allocentric virtual navigation environments. Taken collect ively, while some of our current measures of human spatial ability measure a construct that shares some vari ance with navigation behavior (i.e. CG Arena), th ere are also some important differences between these measures (Moffat, et al., 1997). A third important outcome of the pres ent study was that performance on the CG Arena does not strongly correlate with navigation in real-world space. As with measures of spatial cognition described above, these correlations va ry between the old and young adults but the significant correlation with CG Arena a nd the FRLT delay in the young group suggests that there is reason to suspect that navigation in the real-world is somewhat related to Arena performance. In th e future, and with a larger sample size, it would be useful to follow-up with the exam ination of individual performance in each group to determine if performance in real sp ace is correlated with performance in CG Arena space (i.e. worse/better performers in Arena space predicts better performers in real space). This may help in the interp retation of the varyi ng correlations between groups. While this result is only somewhat consistent with correlational analyses conducted between CG Arena and a real-spa ce analog of the MWM (Laurance et al., 2002), this present study expands on this findi ng even further by evaluating performance on a task that may prove to be more ecologically valid.


42 One criticism of our traditional neuropsychological measures is that they are not particularly effective in predicting ever yday problems (Wilson, 1993). The finding that CG Arena shares some overlapping ability w ith real-world navigation ability suggests that it may be a better predic tor of real-world spatial navi gation impairment than are our traditional tests. Of course, these results are preliminary and ideally would be expanded more fully to real life cognitive mapping envi ronments such as navigating a grocery store or shopping mall. As with the relationship s een with the paper-and pencil measures, the FRLT shares some overlapping spatial ability with the CG Arena, but there is a large amount of variance that appears unique to each measure. Significant group differences (i.e., better performance in the young) on both the immediate and delayed recall of the FRLT are also consistent with previous research that ha s demonstrated age-related decrements in learning routes (Barrash, 1994). In sum, a great deal of interest surrounds the accurate prediction of individuals who will eventually develop AD. This study examin ed, from a clinical perspective, various forms of spatial cognition and navigation in youn g and old adults and their relationship to real world measures of spatial navigation impa irment. It served as an initial step in characterizing particular abilities (e.g., spatial navigation) that might be affected early in the disease process. The next step in this re search will explore the possibility of using the CG arena task as a measure to identify indivi duals at risk for deve loping AD. We plan to test CG Arena in an MCI sample to determ ine if it can effectively predict associated impairment in the medial temporal memory system. The limitations of the current study will be addressed in future work. First, because this study focused on two healthy groups of participants, it was not focused on


43 demonstrating links between behavioral perf ormance and the underlyi ng neural substrate for spatial cognition. Future studies focused on clinical populations will include anatomic measures of MTMS pathology using structur al MRI; demonstrating sensitivity and specificity to anatomic losses would improve th e taskÂ’s clinical and experimental utility. Second, a better measure of real -life allocentric navigation w ould be useful in measuring the CG ArenaÂ’s ecological validity. While the FR LT task used in this study is measure of real life navigation performan ce, it is more a measure of egocentric or route-learning ability. Future comparison with a real world environments requiring more of a map based navigation strategy may be of use (e.g. supe rmarket, shopping mall, see Kirasic, 1991). Third, while our current sample is nearly ma tched on IQ, education, and race, it is not that diverse. Future direction would aim to recruit a more diverse sample that is more representative of the population. In conclusion, the present report provides additional evidence that older adults demonstrate poorer performance on a virtual ta sk of spatial learning and memory than do their younger counterparts. These data confirm th e feasibility of the CG Arena task in an older population and speak to the possible uti lity of using this meas ure in the clinical setting. The present study al so further supports the re lationship between virtual environment navigation and real life navigation. While virtua l allocentric navigation and our current clinical measures of spatial na vigation ability overlap to some degree, CG Arena appears tap abilities that are not othe rwise accounted for with traditional tests and self-report. CG Arena could show promise as a measure to increase sensitivity to prediction of AD and associated impairment in the medial temporal memory system.


44 LIST OF REFERENCES Albert, M. S., Moss, M. B., Tanzi, R., & J ones, K. (2003). Preclinical prediction of AD using neuropsychological tests. Journal of the Internat ional Neuropsychological Society, 7, 631-639. Allen, G. L., Kirasic, K. C., Dobson, S. H ., Long, R. G., & Beck, S. (1996). Predicting environmental learning from spatial abilities: An indirect route. Intelligence, 22 , 327-355. Astur, R.S., Tropp, J., Sava, S., Constable, R.T., & Markus, E.J. (2004). Sex differences and correlations in a virtual Morris wa ter task, a virtual radial arm maze, and mental rotation. Behavioural Brain Research, 151, 103-115. Astur R.S., Taylor, L.B., Mamelak, A.N., Philpott, L., & Sutherland, R.J. (2002). Humans with hippocampus damage display severe spatial memory impairments in a virtual Morris water task. Behavioural Brain Research, 132, 77-84. Astur, R. S., Oritz, M. L., & Sutherland, R. J.(1998). A characterization of performance by men and women in a virtual Morris water task: a large and reliable sex difference. Behav Brain Res, 93 , 185-190. Barrash , Joseph; Damasio, Hanna; & Adolphs, Ralph. (2000). The neuroanatomical correlates of route learning impairment. Neuropsychologia, 38(6), 820-836. Barrash, J. (1994). Age-related de cline in route learning ability. Developmental Neuropsychology, 10, 189-201. Braak, H., & Braak, E. (1991). Neuropathologica l staging of Alzheimer-related changes. Acta Neuropathologica, 82 , 239-259. Bohbot, V.D., Kalina, M., Stepankova, K., Sp ackova, N., Petrides, M., & Nadel. (1998). Spatial memory deficits in patients with le sions to the right hippocampus and to the right parahippocampal cortex. Neuropsychologia, 36(11), 1217-1238. Bondi, M. W., Monsch, A. U., Galasko, D., Butters, N., Salmon, D. P., & Delis, D. C. (1994). Preclinical cognitive markers of dementia of the Alzheimer type. Neuropsychology, 8 , 374-384. Brookmeyer, R., Gray, S., Kawas, C. (1998). Projections of AlzheimerÂ’s disease in the United States and the public health impact of delaying disease onset. American J Public Health, 88, 1337-1342.


45 Bryant, K.J. (1982). Personality correlates of sense of direction and geographic orientation. Journal of Personality and Social Psychology, 43(6), 1318-1324. Cummings, J. E., Tomiyasu, U., Read, S ., & Benson, D. G. (1984). Amnesia with hippocampal lesions after cardiopulmonary arrest. Neurology, 34, 679-681. Dabbs, J. M., Chang, E. L., Strong, R. A., & Milun, R. (1998). Spatial ability, navigation strategy, and geographic know ledge among men and women. Evolution and Human Behavior, 19, 89-98. Driscoll, I., Hamilton, D.A., Yeo, R.A., Brooks, W.M., & Sutherland, R.J. (2005). Virtual navigation in humans: the impact of age, sex, and hormones on place learning. Hormones and Behavior, 47, 326-335. Driscoll, I., Hamilton, D.A., Petropoulos, H., Yeo, R.A., Brooks, W.M., Baumgartner, & Sutherland, R. (2003). The aging hippocampus: Cognitive, biochemical and structural findings. Cerebral Cortex, 13, 1344-1351. Ekstrom, R.B., French, J. W., & Harman, H.H. (1976). Kit of Factor-Referenced Cognitive Tests. Princeton, NJ: Educationa l Testing Service. Flicker, C., Ferris, S. H., & Reisberg, B. (1991). Mild cognitive impairment in the elderly: predictors of dementia. Neurology, 41 , 1006-1009. Folstein, M. F., Folstein, S. E., & Mc Hugh, P. R. (1975). "Mini Mental State." Journal of Psychiatric Research, 12 , 189-198. Frakey, L. (2005). Visuospatial abilities in temporal lobe epilepsy. Retrieved February 4, 2006, from University of FloridaÂ’ s online dissertations website: . Graff-Radford, N.R. (2003). Biological markers. In R. Petersen (Ed.), Mild Cognitive Impairment (pp. 205-227). New York: Oxford University Press. Grundman M., Petersen R., Ferris S., Thomas R., Aisen P., Bennett D., Foster N., Jack C., Galasko D., Doody R., Kaye J., Sano M., Mohs R., Gauthier S., Kim H., Jin S., Schultz A., Schafer K., Mulnard R., Van Dyke C., Mintzer J., Zamrini E., CahnWeiner D., Thal L. (2004). Mild cognitive impairment can be distinguished form Alzheimer disease and normal aging for clinical trials. Arch Neurol, 61, 59-66. Hegarty, M.; Richardson, A. E.; & Montello, D. R. (2002). Development of a self-report measure of environmental spatial ability. Intelligence, 30(5), 425-448. Herbert, L.E., Scherr, P.A., Bienias J.L., Bennett, D.A., & Evans, D.A. (2003). Alzheimer disease in the US population. Arch Neurol, 60, 1119-1122. Hertzog, C. & Rypma, B. (1991). Age differenc es in components of mental-rotation task. Bulletin of the Psychonomic Society, 29, 209-212.


46 Jack, C. R., Jr., Petersen, R. C., Xu, Y. C ., O'Brien, P. C., Smith, G. E., Ivnik, R. J., Boeve, B. F., Waring, S. C., Tangalos, E. G., & Kokmen, E. (1999). Prediction of AD with MRI-based hippocampal volum e in mild cognitive impairment. Neurology, 52 , 1397-1403. Jack, C. R., Jr., Petersen, R. C., Xu, Y. C., Waring, S. C., O'Brien, P. C., Tangalos, E. G., Smith, G. E., Ivnik, R. J., & Kokmen, E. (1997). Medial tem poral atrophy on MRI in normal aging and very mild Alzheimer's disease. Neurology, 49 , 786-794. Jacobs D.M., Marder K., Cote L.J., Sano M., Stern Y., & Mayeux R. (1995). Neuropsychological characteristics of preclin ical dementia in Parkinson's disease. Neurology, 45, 1691-1696. Jacobs, W.J., Thomas, K.G.F., Laurance, H. E., & Nadel, L. (1998). Place learning in virtual space II: Topographic relations as one dimension of stimulus control. Learning and Motivation, 29, 288-308. Jacobs, W.J., Laurance, H.E., & Thomas, K.G.F. (1997). Place learning in virtual space I: Acquisition, overshadowing, and transfer. Learning and Motivation, 28, 521-541. Jacobson M.W., Delis D.C., Bondi M.W., & Salmon D.P. (2002). Do neuropsychological tests detect preclinical Alzheimer' s disease: Individual-test versus cognitive-discrepancy score analyses. Neuropsychology, 16, 132-139. Johnson, K. A., Jones, K., Holman, B. L., Becker, J. A., Spiers, P. A., Satlin, A., & Albert, M. S. (1998). Preclinical predicti on of Alzheimer's disease using SPECT. Neurology, 50, 1563-71. Johnson, K.S. & Albert, M.S. (2003). Func tional imaging. In R. Petersen (Ed.), Mild Cognitive Impairment (pp. 133-148). New York: Oxford University Press. Kato, Y. & Takeuchi, Y. (2003). Individua l differences in wayfinding strategies. Journal of Environmental Psychology, 23, 171-188. King, E.G., Thomas, K.G.F., Bauer, R.M. ( 2005). “Spatial Activities in Environment Questionnaire.” University of Fl orida, Gainesville, Florida. Kirasic, K. C. (2000) Age differences in adu lt’s spatial abilities, l earning environmental layout, and wayfinding behavior. Spatial Cognition and Computation , 2(2), 117134. Kirasic, K. C. (1991). Spatial cognition a nd behavior in young and elderly adults: Implications for learning new environments. Psychology and Aging, 6 , 10-18. Knopman, D.S. (2003). Treatment of mild cognitive impairment and prospects for prevention of Alzheimer's disease. In R. Petersen (Ed.), Mild Cognitive Impairment (pp. 243-258). New York: Oxford University Press.


47 Knopman, D. S., DeKosky, S. T ., Cummings, J. L., Chui, H., Corey-Bloom, J., Relkin, N., Small, G. W., Miller, B., & Stevens, J. C. (2001). Practice parameter: Diagnosis of dementia (an evidence-bas ed review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology, 56 , 1143-1153. Korf, E.S.C., Wahlund, L.O., Visser, P.J., & Sh eltens, P. (2004). Medial temporal lobe atrophy on MRI predicts dementia in patients with mild cognitive impairment. Neurology , 63, 94-100. Krebs, J. R. (1990). Food-storing birds: ad aptive specializat ion in brain and behaviour? Philosophical Transactions of the Roya l Society of London: Series B, 329, 153-160. Krebs, J.R., Sherry, D.F., Healy, S.D., Pe rry, V.H., Vaccarino, A.L. (1989). Hippocampal specialization of f ood-storing birds. Proc Natl Acad Sci USA , 86, 1388-1392. Laurance, H.E., Thomas, K.G.F., Newman, M. C., Kaszniak, A.W., Nadel,L., & Jacobs, W.J. (2002). Older adults map novel envir onments but do not place learn: Findings from a computerized spatial task. Aging Neuropsychology and Cognition , 9(2), 8597. Maguire, E. A., Gadian, D. G., Johnsrude, I. S., Good, C. D., Ashburner, J., Frackowiak, R. S., & Frith, C. D. (2000). Navigation-rela ted structural change in the hippocampi of taxi drivers. Proceedings of the National Academy of Sciences U S A, 97(8), 4398-4403. Maguire, E. A., Burgess, N., & O’Keefe, J. (1999). Human spatial navigation: cognitive maps, sexual dimorphism, and neural substrates. Current Opinion in Neurobiology, 9, 171-177. Mayeux, R., Small, S. A., Tang, M. X., Tycko, B., & Stern, V. (2002). Memory performance in healthy elderly without Al zheimer's disease: effects of time and apolipoprotein-E. Neurobiology of Aging, 22, 683-689. Moffat, S. D., Hampson, E., & Hatzipantelis, M. (1998). Navigation in a “virtual” maze: Sex differences and correlation with psyc hometric measures of spatial ability in humans. Evol Human Behavior, 19, 73-87. Moffat, S.D., Zonderman, A.B., & Renick, S.M. (2001). Age differences in spatial memory in a virtual environment navigation task. Neurobiology of Aging, 22, 787796. Morris, J.C. (2005). Dementia Update 2005. Alzheimer Disease Association Discord , 19(2), 100-117.


48 Morris J., Heyman A., Mohs R.C., Hughes J.P., van Belle G., Fillenbaum G., Mellits E.D., & Clark C. (1989). The Consortium to Establish a Registry for AlzheimerÂ’s disease (CERAD). Part I. Clinical and neuropsychological assessment of AlzheimerÂ’s disease. Neurology, 39, 1159-1165. Morris JC. (1993). The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology, 43, 2412-2414. Morris, R.G., Garrud, P., Rawlins, J.N. & OÂ’ Keefe, J. (1982). Place navigation impaired in rats with hippocampal lesions. Nature, 287 , 681-683. Newman, M. C., & Kaszniak, A. W. (2000). Spatial memory and aging: Performance on a human analog of the Morris Water Maze . Aging, Neuropsychology, and Cognition, 7, 86-93. Nadolne, M. J., & Stringer, A. Y. (2001) . Ecologic validity in neuropsychological assessment: prediction of wayfinding. Journal of the International Neuropsychological Society, 7 , 675-682. OÂ’Keefe, J. & Nadel, L. (1978). The hippocampus as a spatial map . Oxford: Clarendon Press. Oscar-Berman, M. (1994). A comparative ne uropsychology approach to alcoholism and the brain. Alcohol and Alcoholism Supplement, 2, 281-289. Petersen, R.C. & Morris, J.C. (2005). Mild cogn itive impairment as a clinical entity and treatment target. Archives of Neurology, 62, 1160-1163. Petersen, R.C. (2003). Mild cogni tive impairment clinical trials. Nature Reviews Drug Discovery, 2, 646-653. Petersen R. (ed). Mild cognitive impairmen t: Aging to AlzheimerÂ’s disease. 2003; New York: Oxford University Press. Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Tangalos, E. G., & Kokmen, E. (1999). Mild cognitive impairment: clin ical characterization and outcome. Archives of Neurology, 56 , 303-308. Petten, C.V. (2004). Relationship between hi ppocampal volume and memory ability in healthy individuals across the life span: Review and meta-analysis. Neuropsychologia, 42, 1394-1413. Reiman, E. M., Caselli, R. J., Yun, L. S., Chen, K., Bandy, D., Minoshima, S., Thibodeau, S. N., & Osborne, D.(1996). Preclinical evidence of Alzheimer's disease in persons homozygous for the ep silon 4 allele for apolipoprotein E. New England Journal of Medicine 334, 752-758.


49 Roche, R. A. P.; Mangaoang, M. A.; Commin s, S. (2005). Hippocampal contributions to neurocognitive mapping in humans: A new model. Hippocampus , 15(5), 622-641. Salmon, D., & Bondi, M. W. ( 1999). Neuropsychology of Alzheimer disease. In R. D. Terry & R. Katzman & K. L. Bick & S. S. Sisodia (Eds.), Alzheimer Disease (2nd. Ed.) (pp. 39-56). Philadelphia: Lippincott Williams & Wilkins. Sandstrom, N. J., Kaufman, J., & Huettel, S. A.(1998). Males and females use different distal cues in a virtual environment navigation task. Cogn Brain Res, 6, 351-360. Shelton, A.L. & Gabrieli, J.D.E. (2004). Neur al correlates of indivi dual differences in spatial learning strategies. Neuropsychology, 18(3), 442-449. Skelton, R.W., Bukach., Laurance, H.E., Thomas, K.G.F., & Jacobs, W.J. (2000). Humans with traumatic brain injuries show place learning defici ts in a computergenerated virtual space. Journal of Clinical and Ex perimental Psychology, 21, 120. Sherry, D. F., Vaccarino, A. L., Buckenham, K., & Herz, R. S. (1989). The hippocampal complex of food-storing birds. Brain Behavior Evolution, 34(5) , 308-317. Sherry, D.F., Krebs, J.R., Cowie, R.J. (1981). Memory for the location of stored food in marsh tits. Animal Behavior, 29, 1260-1266. Shettleworth, S.J. (2003). Memory and hippocam pal specialization in food-storing birds: Challenges for research on comparative cognition. Brain, Behavior and Evolution, 62, 108-116. Sholl, M.J. (1988). The relationship between sense of direction and mental geographic updating. Intelligence, 12(3), 299-314. Sonnenfeld, J. (1985). Tests of sp atial skill: a validation problem. Man-Environment Systems , 15, 107-120. Squire, L. R. (1992). Declarative and nondeclarative memory: multiple brain systems supporting learning and memory. J. Cogn. Neurosci. 4 , 232-243. Squire, L.R. & Zola, S.M. (1997). Am nesia, memory, and brain systems. Philosophical Transactions of the Royal Society of London B: Biological Science, 352 , 1663-73. Squire, L. R., & Zola-Morgan, S. (1991). Th e medial temporal lobe memory system. Science, 253 ,1380-1386. Sutherland, R. J., Whishaw, I. Q., & Kolb, B. (1983). A behavioural analysis of spatial localization following electrolytic, kainateor colchicine-induced damage to the hippocampal formation in the rat. Behavioural Brain Research, 7, 133-153.


50 Thomas, Kevin G. F.; Hsu, Ming; Laurance, Holly E. (2001). Place learning in virtual space III: Investigation of spatial na vigation training procedures and their application to fMRI and clinical neuropsychology. Behavior Research Methods, Instruments & Computers , 33(1), 21-37. Thomas, K. G. F., Laurance, H. E., Luczak, S. E., & Jacobs, W. J. (1999). Age-related changes in a human cognitive mapping system: Data from a computer-generated arena. CyberPsychology and Behavior, 2 , 545-566. Tierney, M. C., Szalai, J. P., Snow, W. G., Fisher R.H., Nores A., Nadon G., Dunn E., St. George-Hyslop P.H. (1996). Prediction of probable Alzheimer's disease in memory-impaired patients: A prospective longitudinal study. Neurology, 46 , 661665. US Census Bureau. (2002). Census 2000 Summary File 1 (S f1) 100-Percent data, Tables pctl2 and pctl2b. Available at: . Accessed January 30, 2006. Victor, M., & Agamanolis, D. (1990). Am nesia due to lesions confined to the hippocampus: A clinic al-pathologic study. Journal of Cognitive Neuroscience, 2 , 246-257. Voyer, D., Voyer, S., & Bryden, M.P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and considerati on of critical variables. Psychological Bulletin, 117(2), 250-270. Wechsler, D. (1997). WAIS-III, WMS-III technical manual. San Antonio, TX: Psychological Corporation. Welsh, K., Butters, N., Hughes, J., Mohs, R., & Heyman, A. (1991). Detection of abnormal memory decline in mild cases of Alzheimer’s disease using CERAD neuropsychological measures. Arch Neurol, 48, 278-281. Wilson, B. (1993). Ecological validity of neuropsychological assessment: Do neuropsychological indexes predict performance in everyday activities? Applied and Preventive Psychology, 2, 209–215. Wilson, I.A., Ikonen, S., McMahan, R.W., Galla gher, M., Eichenbaum, H., & Tanila, H. (2003). Place cell rigidity co rrelates with impaired spat ial learning in aged rats. Neurobiology of Aging, 24, 297-305. Winblad, B., & Jelic, V. (2004). Long-term tr eatment of Alzheimer's disease: Efficacy and safety of acetylcholinesterase inhibitors. Alzheimer's Disease & Associated Disorders, 18, S2-S8. Zola-Morgan, S., & Squire, L. R. (1986). Memory impairment in monkeys following lesions limited to the hippocampus. Behavioral Neuroscience. 100(2), 155-160.


51 Zola-Morgan, S., Squire, L. R., & Ramus, S. J. (1994). Severity of memory impairment in monkeys as a function of locus and extent of damage within the medial temporal lobe memory system. Hippocampus, 4(4) , 483-95.


52 BIOGRAPHICAL SKETCH Emily Green King graduated with honors fr om the University of Florida in 1998 with a Bachelor of Science. Emily has many years of experience as a psychometrist and research coordinator at Mayo Clinic, Jacksonvi lle, Florida, and a re search assistant and project manger at Beth Israel Deaco ness, Harvard Medical School, Boston, Massachusetts. She entered the clinical a nd health psychology doctoral program at the University of Florida in 2004 with a c oncentration in clinical neuropsychology.